The trend in biomedical imaging is to detect and visualize diseases such as cancer at their earliest time-points, when the disease is most likely to be cured. Molecular imaging techniques are being developed across virtually all imaging modalities for this purpose. Optical molecular imaging offers the potential for not only high imaging resolution, but also multiple approaches for obtaining molecular sensitivity. We have developed an optical molecular imaging technique called Nonlinear Interferometric Vibrational Imaging (NIVI). This technique images the three-dimensional spatial distribution of molecules based on their vibrational resonance frequencies. Molecular vibrational resonances are detected by the nonlinear optical signals from Coherent Anti-Stokes Raman Scattering (CARS). Taking advantage of the coherent nature of CARS, we perform depth-resolved coherence-gating using heterodyne interferometric detection, leveraging many of the principles found in optical coherence tomography (OCT). For the R21 phase of this project, we will extend the imaging capabilities of NIVI to biological molecules associated with cancer. We will improve the sensitivity of detection by rejecting nonresonant signals that are commonly generated from water in biological environments and significantly contribute to background noise, and will implement spectral-domain detection to detect multiple Raman frequencies simultaneously. Before advancing to the R33 phase, we will demonstrate the biological application of NIVI by establishing sensitivity detection limits for biological macromolecules, and by differentiating neoplastic from normal tissue from a rat mammary tumor model based on molecular composition and classification. By demonstrating the potential use of NIVI for biomolecular diagnostics, we will advance this technology toward in vivo imaging in the R33 phase. To accomplish this, we will construct an ultra-broadband NIVI instrument that utilizes femtosecond pulse-shaping techniques to rapidly stimulate multiple vibrational resonances in specific molecules. We will establish the lower sensitivity limits of this technique as well as quantify the resolution at which similar molecular vibrations can be differentiated. Finally to demonstrate the real-time molecular imaging capabilities of our system, we will perform cell, tissue, and in vivo animal studies using a well-characterized carcinogen-induced rat mammary tumor model. Throughout carcinogenesis, tumors will be imaged in vivo to characterize their changing molecular composition and quantify the concentration and spatial distribution of DMA. This research will establish NIVI as a unique in vivo nonlinear optical molecular imaging technique for distinguishing and spatially mapping the distribution of molecules associated with cancer.